Under hidden independent uniform burst points, the 95% interval for early first-pump burst effects included zero for next-trial intentions capped at six in 263 ds000030 records

We tested the effect of a burst at the first pump on next-trial choices in 263 ds000030 version 1.0.0 records. We used nine reward pairs per person. The target was intended pumps capped at six, averaged over all planned pairs. The estimate was -0.247 pumps. A 95% person bootstrap interval was -0.613 to +0.137. It included zero. The analysis did not establish a change or equivalence. Its causal interpretation requires hidden independent uniform burst points from 1 to 12, correct labels, complete potential trials, and source inclusion that preserves assignment. We combined a known first-pump probability with correction of next-trial burst censoring. Exact checks and adaptive simulations passed. A separate parser and calculation matched all pairs and all 20,000 primary bootstrap draws. A full repeat from 266 fresh source files matched all saved analysis files. The package includes proof, reusable code, all source events, hashes, the local plan, results, figures, and repeat instructions. The evidence concerns task behavior during MRI.

Balloon Analogue Risk Task · causal inference · burst censoring · sequential effects · intended pumps · OpenNeuro · ds000030